Market timing

Share this

The advent of ETFs (Exchange Traded Funds) has made the business of trading on the back of so-called 'market timing' signals a lot easier and cheaper than it once was.

Market timing is a distinct subset of technical analysis that has developed as a result of the vast increase in computing power and the ability to sift large amount of stockmarket data quickly and easily.

The way it works is thus.

Take a piece of stockmarket lore. 'Sell in May and go away' is good example. In a US context, the theory that markets generally fall in the first year of a Presidential term might be a good one, or the offbeat (but surprisingly accurate) 'SuperBowl theory', which states that a win for an AFC team presages a stockmarket decline in the following year.

The theory is then backtested to see if it represents a workable trading strategy that generates good long term returns without either isolated disastrous losses ('drawdowns' in the jargon) or significant volatility. Data put into the mix can include economic indicators, other indicators derived from market data such as market yields and PEs, corporate bond yields, moving average patterns and the like.

A lot of both popular and academic work has been done over many years to try devise theories that explain why markets move the way they do, and that link commonly available fundamental and share price data with market movements to try and arrive at some sort of predictive system. Much of this academic work has been converted into formulae and tested against real life experience in the market.

But the 'holy grail' of a perfect predictive theory of stockmarket behaviour is as illusory as ever. Market timing systems work on the basis that if enough subsidiary theories are tested sufficiently rigorously, and the results of the ones generating the best results are aggregated, then the signals thus generated can give a strong (although not infallible) hint as to the future direction of the market.

This is the method adopted by the commercial services that have gone down this route, including Colorado-based Ultra Financial Systems ( and free newsletter services like the Austrian SNIPER system ( Systems like this deal in probabilities rather than certainties, as they themselves are quick to make clear. But both of the examples, the proprietors to trade in their own accounts exclusively on the back of the signals their systems generate.

Market timing systems like this and other similar ones tend to categorise their backtested system theories, and the signals they generate, into those applying to stocks (split between short-term and medium-long term), bonds and gold. Ultra focuses primarily on US markets: SNIPER looks at a wider range of global markets providing 'buy or hold' or 'move to cash' signals for all of them, as well as an overall risk indicator that measures sentiment in all markets covered.

Both reject many more systems than they include after going through the backtesting process. In the case of Ultra, for example, the 100 or so indicators in his company's system are the end result of a process that finds 80% of theories tested being rejected as unreliable.

It is probably worth stressing that some of the indicators included in systems like this produce modest results and rely on frequent trading to generate acceptable returns, while others rely on spotting longer term secular trends. Either way, profits have to be maximised by keeping down costs, which means that actively traded ETFs, where spreads are tight, and transaction costs low, are the ideal vehicle to employ. For those using techniques like this, it is probably as well to avoid introducing leverage of any description into the equation.

There is also a large question mark over whether negative signals from market timing systems should be taken as an invitation to 'short' the market. The testing of theories that underlie a market timing system usually assume that if a theory is not suggesting a 'long' position, the alternative is simply to be in cash, and it is wise move simply to observe this rule without question.